How a Two-Person Newsletter Crew Used DiggHits to Catch a Breaking Story Before the Big Outlets

We noticed something odd in our inbox last month: a small newsletter we follow, run by two pseudonymous editors who call themselves Plex and Marrow, broke a story about a municipal broadband outage a full 90 minutes before any major outlet. They credited a social news aggregator we'd been meaning to look at for months. So we followed the project from tip to publish to see whether the outcome was luck or process.

The tool in question was DiggHits, a community-ranked feed where every submission is voted up or down by real humans rather than sorted by an opaque recommendation engine. Plex described it as "the front page edited by the crowd, not by a black box." That framing matters for a two-person operation with no budget for wire services.

The setup: one dashboard, two editors, and a 6 a.m. scan

Plex and Marrow run a five-day-a-week briefing on local infrastructure and consumer tech. Before this experiment, their morning routine was a mess: 14 open browser tabs, three RSS readers, and a Slack channel full of half-read links. Their problem wasn't access to information. It was triage.

On day one, they built a simple workflow. Each editor spent 20 minutes scanning the rising section of the aggregator, upvoting anything that touched broadband, utilities, or platform policy. They kept a shared note of vote counts at three checkpoints: 8 a.m., noon, and 4 p.m. No automation, no paid tier, just discipline and a stopwatch.

The first week produced nothing dramatic. The second week, the pattern shifted. A submission about a regional ISP quietly throttling upload speeds climbed from 40 votes to more than 600 in under three hours. That velocity, not the raw number, was the signal. "A story that jumps that fast on a Tuesday morning is either astroturfed or real," Marrow told us. "We called the utility. It was real."

Decision points: when to trust the crowd and when to verify

The team's rule was blunt: the aggregator could suggest a story, but it could never confirm one. Every high-velocity item got a 15-minute verification window — one phone call, one public records check, or one on-the-record email. If two of three checks failed, the item was dropped, regardless of how many votes it had.

• Threshold one: 200 votes within two hours triggered a look.
• Threshold two: 500 votes plus a comment thread with primary sources triggered a call.
• Threshold three: any submission cross-posted by three or more unrelated accounts triggered a manual review for coordinated voting.

The obstacle they hadn't anticipated was noise during breaking events. When the outage hit, the feed filled with duplicate submissions and speculation. Plex and Marrow had to ignore the vote count entirely and read the comment threads instead, looking for residents posting screenshots and timestamps. That human filtering is exactly the point of a community-ranked system, but it still requires a human to do the filtering.

We asked whether they ever worried about being manipulated. Marrow laughed. "Every platform can be gamed. The difference is whether you can see the game happening." On a transparent feed, vote patterns and account histories are visible. On an algorithmic feed, they aren't. That visibility is what let the pair spot a coordinated push on a rival story and set it aside.

Measurable results: 90 minutes, 4,100 readers, one correction

The payoff arrived on a Thursday. A submission about a fiber cut affecting three counties hit 800 votes in 71 minutes. Plex made two calls, got a confirmation from a county spokesperson, and published at 9:14 a.m. The first major outlet followed at 10:44 a.m. Their post drew 4,100 readers — modest by national standards, enormous for a list of 6,000 subscribers. Open rate that day was 61 percent, well above their 34 percent average.

They also issued one correction, which they consider a success rather than a failure. An early vote spike pointed to a second outage that turned out to be a scheduled maintenance window. Because they had logged the vote trajectory, they could trace the error to a single misread comment and fix it within 40 minutes. DiggHits reports 1.8M+ monthly active diggers across 142 countries, with 47M votes cast in the last year, so the volume is real — and so is the need for editorial judgment on top of it.

The broader lesson for makers and small publishers is unglamorous. A community-ranked feed doesn't replace reporting. It shortens the distance between a signal and a decision. For a two-person crew, that distance was the whole bottleneck.

If you want to see how the ranking and voting actually work before building your own routine, the platform's own explanation of its submission and moderation mechanics is worth 10 minutes. We'd suggest starting with the breakdown of how community voting shapes the front page, then running the three-threshold test above for two weeks. Track your hit rate. Ours improved from one in nine to one in four, and we're still tinkering.